A model of unitary responses from A/C and PP synapses in CA3 pyramidal cells (Baker et al. 2010)
ModelDBRepository/137259
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <meta content="text/html; charset=ISO-8859-1" http-equiv="Content-Type"> <title>Neuron model for Baker et al. 2010 JCNS</title> </head> <body> This folder/directory contains files supporting Neuron simulations described in:<br> <br> Baker JL, Perez-Rosello T, Migliore M, Barrionuevo G, Ascoli GA (2010). A computer model of unitary responses from associational/commissural and perforant path synapses in hippocampal CA3 pyramidal cells. J Comput Neurosci (epub ahead of print). <br> <br> <big style="font-weight: bold;">Overview</big><br> <p>Unlike some Neuron models, the intent of this model is primarily to facilitate batch simulations involving the analysis of thousands of individual data traces across multiple cells. Data traces from single synaptic simulations can be obtained as described below, but no user-friendly pushbutton interface is provided for doing so. Batch simulations create summary results for each synaptic activation but do not save individual data trace results.</p> <p> In executing the model it may be useful to understand the role of the different HOC files provided. These files fall into the following groups. </p> <ol> <li>HOC files that define the geometry of individual cells. For example,<span style="font-family: monospace;"> ama-c30573.CNG.hoc</span> describes cell C30573 from the Amaral lab. File <span style="font-family: monospace;">axon-common.hoc</span> describes a common axon segment that is attached to each of the individual cells during the course of the simulation. Note that these geometries were generated from <a href="http://NeuroMorpho.org">NeuroMorpho.org</a> and hence may differ in detail from similar descriptions of the same cells found elsewhere.<br> <br> </li> <li>HOC files that provide parameters specific to individual cells. These parameters account for spatial orientation of the apical direction of the cell and the division between layers stratum oriens (SO), stratum lucidium (SL), stratum radiatum (SR), and stratum lacunosum-moleculare (LM). The output file name for simulations of synaptic responses throughout the cell is also supplied.<br> </li> <li>HOC file <span style="font-family: monospace;">synresp.hoc</span> contains both the common parameters of the simulation and the logic for actually running the simulation. File <span style="font-family: monospace;">synresp.hoc</span> is loaded by the HOC file containing parameters for an individual cell, ensuring that cell-specific parameters have already been set when <span style="font-family: monospace;">synresp.hoc</span> is loaded. Only one cell at a time is simulated per execution of the model. </li> </ol> <big><span style="font-weight: bold;">Installation</span></big><br> <p>Files from this model can be copied to a directory of the user's choice. The model can then be executed using a recent release of Neuron. Versions 6 and 7 were used during the preparation of the article, but there should be only small differences between results using different execution platforms and different Neuron releases. Of course, MOD files must be compiled before the model can be executed. This is done using the usual commands. For MS Windows use the <span style="font-family: monospace;">mknrndll</span> command and for Unix use<span style="font-family: monospace;"> nrnivmodl</span>. File <span style="font-family: monospace;">exp2nmdar.mod</span> is used to simulate NMDA receptors using a dual-exponent model. Other MOD files are associated with the various active models considered in the article.</p> <p> Files <span style="font-family: monospace;">demo.hoc</span>, file <span style="font-family: monospace;">demo.ses,</span> and files beginning with <span style="font-family: monospace;">demo-fig2a</span> are included to permit a simple verification of the installation of the model and to illustrate a method for obtaining the results of a single synaptic activation by invoking functions from the Neuron command line. This demonstration reproduces the synaptic activation shown in Figure 2a of the article. It can be launched either directly from <span style="font-family: monospace;">demo.hoc</span> or else by loading <span style="font-family: monospace;">demo.hoc</span> into Neuron using the <span style="font-family: monospace;">load hoc</span> menu item, provided that the current directory has previously been set to the directory containing the files of this model. <span style="font-family: monospace;">demo.hoc</span> automatically loads the other files as needed in turn. Results of executing the demo<span style="font-family: monospace;"></span> should, after some rearranging of windows, look like the following.<br> </p> <p><img style="width: 800px; height: 514px;" alt="Neuron windows from demo.hoc" src="demo.png"><br> </p> <big><span style="font-weight: bold;">Execution</span></big><br> <p>Simulation parameters are defined in the beginning of file <span style="font-family: monospace;">synresp.hoc</span> and must be set as appropriate for the type of simulation being performed. See <span style="font-family: monospace;">synresp.hoc</span> for a description of each of these parameters. For parameter settings used in generating files supporting figures in the article, see file <span style="font-family: monospace;">params-by-fig.csv</span>. Note that <span style="font-family: monospace;">synresp.hoc</span> is loaded by one of the cell-specific HOC files and will not function correctly without the necessary cell-specific parameter values being set before <span style="font-family: monospace;">synresp.hoc</span> is loaded. </p> <p> Simulations are started by invoking Neuron and loading one of the cell-specific HOC files. This can be done in the usual way, but depends on the type of system being used. For example, in MS Windows, Neuron is typically associated with files containing the suffix <span style="font-family: monospace;">.hoc</span> and it is sufficient to double-click on a cell-specific HOC file (for example <span style="font-family: monospace;">synresp-cell1zr.hoc</span>) to launch the simulation. Before loading the cell-specific HOC file using the Neuron menu item <span style="font-family: monospace;">File->load hoc</span>, set the current directory to be the directory containing <span style="font-family: monospace;">synresp.hoc</span> and other model files using the menu item <span style="font-family: monospace;">File->working dir</span>.</p> <p> Control parameters of particular interest are <span style="font-family: monospace;">isInteractive</span>, which controls whether interactive components of Neuron are loaded, and <span style="font-family: monospace;">runStim</span>, which controls whether or not stimulations for synaptic activations are initiated automatically when <span style="font-family: monospace;">synresp.hoc</span> is loaded. For batch executions, normal settings would be <span style="font-family: monospace;">isInteractive=0</span> and <span style="font-family: monospace;">runStim=1</span>. For running simulations under manual control, the settings would be <span style="font-family: monospace;">isInteractive=1</span> and <span style="font-family: monospace;">runStim=0</span>, which are the values set initially in the copy of <span style="font-family: monospace;">synresp.hoc</span> supplied here.</p> <span style="font-style: italic;">Manually Controlled Execution</span><br> <p>When executing manually controlled simulations, the following functions defined in synresp.hoc can be invoked from the Neuron command line: </p> <ul> <li><span style="font-family: monospace;">getSynResp(x)</span> simulates a synaptic activation at location x within the current section. Summary results are written to the Neuron console log. Normal Neuron plotting mechanisms can be used to display state variables of relevant portions of the cell, but the required plotting settings would need to be put into place prior to invoking <span style="font-family: monospace;">getSynResp</span>. Note that if the current section and location lie at the soma or in layer SL, there is no corresponding synapse. In this case, the function returns without conducting a simulated activation.</li> <li><span style="font-family: monospace;">saveSynResp(x)</span> simulates a synaptic activation at location x within the current section. Results are written to the file named in string variable savePath (default is<span style="font-family: monospace;"> savedsynresp.csv</span>). </li> </ul> <p>Output files are in comma-separated-values (<span style="font-family: monospace;">.csv</span>) format and contain the following columns:</p> <table style="text-align: left; width: 810px; height: 172px;" border="0" cellpadding="2" cellspacing="2"> <tbody> <tr> <td><span style="text-decoration: underline;">Column</span><br> </td> <td><span style="text-decoration: underline;">Contents</span><br> </td> </tr> <tr> <td style="text-align: left; font-family: monospace;">time</td> <td> event time in ms</td> </tr> <tr> <td style="text-align: left; font-family: monospace;">soma</td> <td> either somatic current injection for voltage-clamp (pA) or somatic membrane potential for current-clamp (mV)</td> </tr> <tr> <td style="text-align: left; font-family: monospace;">dend</td> <td> membrane potential in the dendrite at the site of the synapse (mV)</td> </tr> <tr> <td style="font-family: monospace;">iampar<br> </td> <td>AMPA receptor currents (pA)<br> </td> </tr> <tr> <td style="text-align: left; font-family: monospace;"> inmdar</td> <td>NMDA receptor currents (pA)</td> </tr> </tbody> </table> <p>Sample data traces used in the article are included here to permit testing of any new installations of the model. Correspondence between figures and the traces files is:</p> <table style="text-align: left; width: 810px; height: 200px;" border="0" cellpadding="2" cellspacing="2"> <tbody> <tr> <td style="vertical-align: top;"><span style="text-decoration: underline;">Figure</span><br> </td> <td style="vertical-align: top;"><span style="text-decoration: underline;">Sample Trace File Name</span><br> </td> </tr> <tr> <td>2a<br> </td> <td style="font-family: monospace;">out-vc-ampar-c31162-ad67-022.csv</td> </tr> <tr> <td>2b<br> </td> <td style="font-family: monospace;">out-vc-ampar-c62563-ad54-054.csv</td> </tr> <tr> <td>2c<br> </td> <td style="font-family: monospace;">out-vc-nmdar-l51-ad7-036.csv</td> </tr> <tr> <td>2d<br> </td> <td style="font-family: monospace;">out-vc-nmdar-c81463-ad87-082.csv</td> </tr> <tr> <td>8a SR<br> </td> <td style="font-family: monospace;">out-vc-ampar-c62563-ad2-01667.csv</td> </tr> <tr> <td>8a SLM<br> </td> <td style="font-family: monospace;">out-vc-fastampar-c62563-ad2-01667.csv</td> </tr> </tbody> </table> <p>Note that the synaptic stimulation event occurs at t=2000 ms. Data values appearing prior to that time are associated with settling of the simulated cell from its initial rest state under the influence of either a current or voltage clamp at the soma as appropriate. After loading the model into Neuron, data traces were written to external files using HOC commands of the form:</p> <div style="margin-left: 40px;"><span style="font-family: monospace;">apical_dendrite[n]{saveSynResp(x)}</span><br> </div> <br> where <span style="font-family: monospace;">n</span> is the apical dendrite section number and <span style="font-family: monospace;">x</span> is the position within the identified section. <p>By convention, values for the receptor type, clamp mode, cell name, section number, and position were encoded into the output file name, though this is not enforced by the model software. The output file name should be placed in string variable <span style="font-family: monospace;">savePath</span> before using the above command. <br> </p> <p>Data in the generated files can be plotted using a number of tools. Most statistical packages are capable of reading .csv files such as these. For example, when using the statistical package <a href="http://www.r-project.org">R</a>, the following procedure can be used to generate a simple EPSC plot: </p> <ol> <li>In an R workspace, set the current directory to one containing the trace files.</li> <li>Execute the following commands in the workspace:<span style="font-family: monospace;"><br> <br> SR<-read.csv("out-vc-ampar-c31162-ad67-022.csv")</span><span style="font-family: monospace;"><br> SR<-SR[SR$time>=1990 & SR$time<=2060,]</span><span style="font-family: monospace;"><br> plot(SR$time-2000, # offset by settling time<br> SR$soma-SR$soma[1], # offset by resting current<br> type='l',xlab="time",ylab="EPSC")</span><span style="font-family: monospace;"><br> rm(SR)</span> <br> </li> </ol> <span style="font-style: italic;">Batch Execution</span><br> <p>When <span style="font-family: monospace;">runStim=1</span> is specified, the function <span style="font-family: monospace;">sampleSynResps()</span> is automatically invoked to sample synaptic responses throughout the current cell. A comma-separated-values (<span style="font-family: monospace;">.csv</span>) file is written containing the following columns:</p> <table style="text-align: left; width: 100%;" border="0" cellpadding="2" cellspacing="2"> <tbody> <tr> <td style="vertical-align: top;"><span style="text-decoration: underline;">Column</span><br> </td> <td style="vertical-align: top;"><span style="text-decoration: underline;">Contents</span><br> </td> </tr> <tr> <td style="font-family: monospace;">section<br> </td> <td>section name of the section containing the synapse being simulated (string)<br> </td> </tr> <tr> <td style="font-family: monospace;">loc<br> </td> <td>location of the synapse within the section containing the synapse (value 0-1)<br> </td> </tr> <tr> <td style="font-family: monospace;">layer<br> </td> <td>layer containing the synapse. One of: SO, SR, or LM.<br> </td> </tr> <tr> <td style="font-family: monospace;">y<br> </td> <td>synapse y-axis value along the reoriented cell apical direction (microns, SO negative)<br> </td> </tr> <tr> <td style="font-family: monospace;">dist<br> </td> <td>path distance from the cell soma midpoint to the synapse (microns)<br> </td> </tr> <tr> <td style="font-family: monospace;">len<br> </td> <td>axial length of the segment containing the synapse (microns)<br> </td> </tr> <tr> <td style="font-family: monospace;">area<br> </td> <td>membrane area of the segment containing the synapse (microns)<br> </td> </tr> <tr> <td style="font-family: monospace;">trunk<br> </td> <td>1 if the synapse lies on a contiguous path from the soma to layer LM; 0 otherwise<br> </td> </tr> <tr> <td style="font-family: monospace;">type<br> </td> <td>recording mode simulated. CC for current-clamp and VC for voltage-clamp<br> </td> </tr> <tr> <td style="font-family: monospace;">peakValue<br> </td> <td>peak somatic response relative to rest (pA for VC, mV for CC)<br> </td> </tr> <tr> <td style="font-family: monospace;">peakTime<br> </td> <td>time of peak somatic response following synaptic activation (ms)<br> </td> </tr> <tr> <td style="font-family: monospace;">halfWidth<br> </td> <td>half-height width of somatic response (ms)<br> </td> </tr> <tr> <td style="font-family: monospace;">synRest<br> </td> <td>rest membrane potential at the location of the synapse (mV)<br> </td> </tr> <tr> <td style="font-family: monospace;">synPeak<br> </td> <td>peak membrane depolarization at the synapse (mV, not relative to rest)<br> </td> </tr> <tr> <td style="font-family: monospace;">gAMPAR<br> </td> <td>conductivity of simulated AMPA receptor (nS)<br> </td> </tr> <tr> <td style="font-family: monospace;">tau1AMPAR<br> </td> <td>fast time constant of AMPA receptor model (ms)<br> </td> </tr> <tr> <td style="font-family: monospace;">tau2AMPAR<br> </td> <td>slow time constant of AMPA receptor model (ms)<br> </td> </tr> <tr> <td style="font-family: monospace;">gNMDAR<br> </td> <td>conductivity of simulated NMDA receptor (nS)<br> </td> </tr> <tr> <td style="font-family: monospace;">tau1NMDAR<br> </td> <td>fast time constant of NMDA receptor model (ms)<br> </td> </tr> <tr> <td style="font-family: monospace;">tau2NMDAR<br> </td> <td>slow time constant of NMDA receptor model (ms)<br> </td> </tr> </tbody> </table> <br> <p>The path name of the output file must be placed in the variable <span style="font-family: monospace;">outPath</span>. This is done in the cell-specific HOC files before <span style="font-family: monospace;">synresp.hoc</span> is loaded. Note that for batch execution, individual synaptic activation data traces are not saved. </p> <p> Procedures for automating simulations involving multiple cells can be implemented in Unix systems through shell scripts and in MS Windows through batch command files. In this case, it may be useful to redirect<span style="font-family: monospace;"> sysout</span> to a file so that the simulation results can be scanned for any error messages. Similarly, scripts can be developed to improve load balancing when simulating multiple cells on a multi-core system. These, however are left as an exercise for the reader because of differences among execution platforms.</p> <big style="font-weight: bold;">Author Contact Information</big><br> <p>For questions regarding this model, please contact the article's corresponding author, Dr. Giorgio Ascoli. His email address is (replace -at- with the usual @ symbol)<span style="font-family: monospace;">: ascoli-at-gmu.edu</span> </p> <br> <br> </body> </html>
About
A model of unitary responses from A/C and PP synapses in CA3 pyramidal cells (Baker et al. 2010)
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published